Download The Role of Theory in Educational research (2011)

The Role of Theory
in Educational Research
Report from the March Seminar 2011
Programme
Norwegian Educational Research towards 2020 - UTDANNING2020
About the programme
Norwegian Educational Research towards 2020 - UTDANNING2020
The Programme for Educational Research – UTDANNING2020
(2009–2018) succeeds the Programme for Knowledge, Education and Learning – KUL (2003–2007) and runs parallel with
the Programme for Practice-based Educational Research –
PRAKUT (2010–2014). The programme has a total budgetary
framework of approximately NOK 400 million.
The UTDANNING2020 programme conducts research on the
education sector – from early childhood education and care
to doctoral level education. The programme seeks to enhance
the knowledge base for policymaking, public administration, professional education and professional practice and to
promote research of high scientific merit. A variety of subject
areas and research communities are encouraged to conduct
research on issues related to the education sector as well as
areas of overlap in other sectors.
Programme Board
Contents
Kirsti Klette, Professor, University of Oslo (chair)
Lars-Erik Borge, Professor, Norwegian University of Science and
Technology
Mikael Alexandersson, Professor, University of Gothenburg
Knud Knudsen, Professor, University of Stavanger
Eivind Heder, Director General, Ministry of Education and
Research
Jorunn Dahl Norgård, Senior Executive Officer, Union of
Education Norway
Jorun Sandsmark, Adviser, Norwegian Association of Local and
Regional Authorities (KS)
Ingrid Helgøy, Research Leader, Stein Rokkan Centre for Social
Studies, University of Bergen, (deputy member)
Ingegerd Tallberg Broman, Professor, Malmö University
(deputy member)
Observer from the PRAKUT programme
Kirsti Klette:
The Role of Theory in Educational Research . . . . . . . . . . . . . . page 3
Agnes Tellings:
Theories and Research in the Field of Education:
An Indissoluble Union . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . page 8
Lars Mjøset:
Many Notions of Theory – Too Few Methodologies
to Deal with Them . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . page 16
Siv Haugan:
Multifaceted or one-dimensional? . . . . . . . . . . . . . . . . . . . . . . page 26
Introduction
The Role of Theory in Educational Research
As a means of promoting scientific quality and merits, UTDANNING2020 uses different tools
and strategic actions. A special annual seminar – held in March every year – is devoted to addressing scientific and methodological issues in educational research. The first March seminar
took place in 2010, and discussed rigour and relevance in educational research. Last year the
seminar was devoted to “The Role of Theory in Educational Research”. Scholars from different
social sciences devoted to educational research (educationalists, economists, political scientists, sociologist and so) shared their insight and reflections on the role and use of theory in
educational research. This publication reflects some of the contributions at this seminar.
The notion and term of theory are essential in any discipline
that perceives itself as scholarly or scientific. Suppes (1974: 2)
claimed already forty years ago that the obvious and universally recognised importance of theory in the more mature
sciences is strong evidence of the universal generalisation
that theory is important in all sciences. Hence, theory is essential in educational research as a research domain, and the
term is frequently used in papers, books, and – not least – in
PhD. dissertations. On closer inspection, however, it is not at
all clear what theory actually means in the different social
sciences devoted to educational research. Nor is it clear where
the entities referred to as theories invoked in educational
studies come from, how they are used, what foundations they
have, or what roles they play in the field. It seems problematic
that a key entity for the advancement of research in a field is
ill-defined and has an unclear status and function. The report
you are now reading is therefore an attempt to help put these
issues on the agenda of articulate and deliberate discussions
within our field of expertise. The contributions in this report
reflect different voices heard at the 2011 March seminar
devoted to the role of theory within the educational sciences.
This report is not intended to survey or review in a coherent
way different specific theories put to use in the educational
sciences. Rather, it is meant to offer some general challenges
and considerations of an overarching nature regarding the
role of theory within this field of research.
So what is theory in the scientific field of educational
research?
Simply speaking, theory refers to a particular kind of explanation. Leedy and Ormrod (2005: 4) state: “A theory is an organized body of concepts and principles intended to explain a
particular phenomenon”. Thus, theories explain how and why
something functions the way it does (Johnson & Christensen,
2007: 7). As pointed out by Boss, Doherty, LaRossa, Schumm, &
Steinmetz (1993: 20): “Theorizing is the process of systematically formulating and organizing ideas to understand a particular phenomenon. Thus, a theory is the set of interconnected
ideas that emerge from this process”. Following McMillan and
Schumacher (2000), a theory can develop scientific knowledge
33
congruent with the following criteria: first, provide simple
explanation about the observed relations regarding their relation to a phenomenon; second, be consistent with an already
founded body of knowledge and the observed relations; third,
provide a device for verification and revision; and fourth, stimulate further research in areas in need of investigation.
So in principal, for a system of concepts and claims to be
called a theory, the system has to be i) stable, i.e. unchanged
over a longer cycle of time, ii) coherent, i.e. the components of
the system have to be linked in a comprehensive and non-contradictory way, and iii) consistent in the sense that it should
not be possible to arrive at contradictory claims by means of
the types of derivation permitted in the theory. In practice,
however, many systems named theories do not possess all
these features. This is particularly true of theories in educational sciences the mathematical educationalist Niss (2006)
argues. Some theorists would add that the requirements that
all non-fundamental claims in a theory have to be testable. In
other words, such theorists would reject what Niss describes
as “transcendental theories” (2006: 3), i.e. theories in which
the concepts and claims are so general and overarching that
they do not apply in a straightforward way to a specific,
empirically well-defined world. A number of theories belonging to the humanities or the social sciences and education are
transcendental in this sense.
Once social/educational sciences are required to hold the
same features as natural sciences in terms of their theories,
it becomes evident that social sciences have more competing
theories available. Anfara and Mertz (2006) declare that the
major characteristic of social sciences in contrast to natural
4
sciences is its multiple theoretical orientations which never
reaches a fixed consensus like the empirical referents or
explanatory schemes which characterise natural sciences. The
competing theories available in social sciences are commonly
popular because the nature of the phenomenon that is being
studied allows its consideration through multiple perspectives and thus each of these perspectives suggests a reasonable explanation of the phenomenon.
The many faces and functions of theory in educational
research
As outlined above theory has many functions within the
educational sciences , just as it does in the social sciences.
Theories provide predictions and explanations as well as
guidelines for actions and behaviour. Theories might provide
a structured set of lenses through which aspects or parts
of the world can be observed, studied or analysed. Theories
also provide a safeguard against unscientific approaches to
a problem, an issue or a theme. By articulating underlying
assumptions and choices and by making them explicit and
subject to discussion and by situating one’s research within
some theoretical framework one might be protected towards
criticism. In his insightful paper on “The Place of Theory in
Educational Research” dating back to 1974, Suppes identified
five different ways in which theory might be made relevant
for educational research by analogy, by reorganizing of experiences, as a device for recognizing complexity, for problem
solving and to avert the triviality of “empiricism”.
The first, theory as a means of building analogies and universal generalisations, goes back to how knowledge is built in the
more mature sciences like physics and other natural sciences.
This is the classical and standard notion of theory Mjøset argues for in this publication, and it serves as a reservoir to make
predictions on the basis of analogue situations. The second,
theory understood as the reorganising of experiences, makes
theory relevant as it can change our perspective on what is important and what is superficial. For many cases where theory
has been successful in science we can make a strong argument
for the deeper organisation of experience which the theory
has thereby provided. The third way posits that theory also
nurtures our ability for recognizing complexity. One benefit of
theory is to show that what appears on the surface to be simple matters of empirical investigation, on a deeper level, turns
out to be complex and subtle. It is a merit of theory to push for
a deeper understanding of the acquisition and not to relax before we have a complete analysis of what a student does and
what goes on inside his/her head as (s)he for example acquires
a new skill. The fourth notion, problem solving as theory, goes
back to the Deweyan concept of inquiry as the building block
and repository for the transformation of an indeterminate and
indefinite situation that presents a problem regarding what is
determinate and definite. Inquiry - and problem solving- as a
theoretical device can thus be seen as corrective to the overly
scholastic and rigid standard conception of scientific theory,
and where the developments of scientific theories are regarded as potential tools for problem solving. Suppes’ fifth way is
where theory serves as a protection against the triviality of
“empiricism”. Those parts of science that have been beset by
“empiricism” have been suffering accordingly. The deficiency
of simple empirical facts and anecdotes is well acknowledged.
Empiricism as simply the recordings of individual facts and
with no apparatus of generalisation or theoretical framework
leads nowhere.
As indicated in this short overview, theories serve many functions and have many features in the social/educational sciences,
and the variability within each of these features of a theory is
sufficient to suggest that the notion of theory is not exactly a
unified one. This poses severe challenges for the role of theory
in the field of educational research. In the next section I will end
this discussion by postulating four general challenges regarding
the role of theory in the educational sciences.
Challenges in the use of theories in the educational sciences
First, there is no such thing as a well-established unified “theory
of education” which is supported by the majority of educational researchers. On the contrary, different groups of researchers
represent different schools of thought, some of which appear
to be mutually incompatible if not directly contradictory. Moreover, for reasons that will be discussed later in this report it is
neither likely nor desirable that we shall get a unified theory in
educational research in a foreseeable future, if ever.
Secondly, many educational researchers relate their work to
explicitly invoked theories borrowed from other fields (or at
least from authors who belong to other fields), and often do
so in rather eclectic or vague ways. The use of ‘grand’ theory,
or theories about the social world as advocated by such major
social theorists as Karl Marx or Michel Foucault, for example,
raises further and different issues for research engagement.
When basing research in the understandings of a specific
theorist or set of thinkers it can be tempting to apply the
theory in a more or less uncritical fashion to interpretation of
evidence, rather than engaging critically with the theory itself
or engaging with the evidence in an effort to test or modify
the theory. At its worst, this can resemble the “adulation of
5
great thinkers” (Tooley & Darby, 1998) or what the sociologist
Frønes in this report entitles “fashion of the nonsense”. Rather
seldom are theories “homegrown” (Niss, 2006: 4) within the
educational sciences.
Thirdly, much discussion and debate within educational research takes the shape of “battles” with and between theories.
This could potentially be fruitful to the degree that competing theories offer different perspectives on the same thing,
whereas it is potentially unproductive, if not damaging, if the
theories deal with different things and therefore compete only
in the superficial sense that “my topic object of study is more
important than yours”. In this report Tellings suggests different
mechanisms of synthesising as a means of going beyond what
appear to be competing theoretical perspectives.
Fourth and finally, quite a few educational researchers have a
poor match between the theory they invoke and its relevance
for their data set .Many researchers who actually do invoke
a theory in their publications do not seem to go beyond the
mere invocation, and with subsequently a poor fit in applying the theory to their data and the interpretations of their
evidence. In other words, some theoretical framework may be
referred to in the beginning or in the end of a paper without
having any presence or bearing on what happens between
the beginning and the end.
The three contributions that follow engage with the role
of theory in different ways. Tellings is occupied with the
many functions and faces of theory in educational research.
Theories might both describe and explain reality, and at its
best also change reality. As for the latter she argues that it is
likely that we need to combine different theoretical positions. For example, in order to learn fractions in mathematics you have to combine insight from language learning and
conceptual knowledge, but also real-life experience with
quantities. Furthermore, you have to draw on short-term
and long-term memory, rote learning, and insight into how
6
fractions are solved. A good theory concerning fractions
must account for all of this, she argues, and it is likely that
a combination of different theories will work best. Tellings
(2001) is subsequently occupied with integrative efforts in
the educational sciences, an attempt she sees as yet to be
made. Tellings suggests four different mechanisms of integration: reduction, synthesis, horizontal addition, and vertical
addition. For the domains of education with it multidisciplinary character, horizontal addition is especially promising
and powerful, she argues.
From the point of social science sociologist Lars Mjøset underscores natural sciences – what he entitles standard view of science, social philosophy (humanities, linguistics, philosophy and
other text related disciplines) and a contextualised view (case
studies) as three distinct theoretical traditions. He argues that
each of these views connects to a reference faculty within the
university structure and with definite traditions, history and
methodological guidelines. The social-philosophical reconstructive notions of theory build on Kantian ideas of enlightenment
and encompass transcendental reasoning such as philology,
case constructions, and other text-based ways of thinking.
Today these ideas are recognized in contemporary European
philosophy within “grand theories” such as Luhmann, Giddens,
and Habermas, Mjøset states.
The standard law oriented notion of theory builds on the
natural sciences such as physics experiments and other natural sciences. The tradition has contributed to foregrounded
experimental and statistical design inspired theories within
the educational sciences. Most social sciences today work with
such data, regarding statistical inference as quasi experiments,
reading regression equations as evidence, and econometrics.
Mjøset sees contextualize theories as the most promising
and adequate for the educational sciences, and especially the
contextualized explanation view of theory. The contextualized explanation view of theory has its origins in ethnography,
anthropology and other cased based, and bottom up, based
models for generating theory, and which retain contexts and
achieve generalization that are limited in time and space. Contextualize theories are not a method community –theories
build from explanation of cases with reference to selected
case categories, Mjøset argues. Like Tellings, he sees crosscase analyses and similar integrative efforts as a means by
which contextualized explanation theories could escape away
from their focus on unique cases and singular units.
The last contribution in this report summarises the roundtable
debate on educational theories as a field of cross-disciplinary
investigation. From the disciplines of education, political science, economics and sociology, Haug, Frølich, Bonesrønning
and Frønes discuss the role of theory within the education sciences. Frønes questions the tendency to privilege transcendental theories (Niss, 2006: 3) and grand theories, what he labels
“fashion of the nonsense”. Haug is sceptical towards what he
sees as a tendency towards privileging certain theoretical positions and “one theory fits all”. All four of them, however, underscore educational research as a multidisciplinary field where
different theories should work in concert. They also underscore
educational research as a multifaceted field of investigation
where a magnitude of theoretical positions is required and
desirable. There is no single theoretical position that could
serve all purposes. Quite the contrary, they argue, which is why
we need all kinds of “workers in the vineyard”, and we need a
variety of training for these various workers, not only in terms
of different theories of education, but also in terms of whether
their approach is primarily theoretical or empirical.
References:
Anfara, V. A., & Mertz, N. T. (Eds.) (2006). Theoretical Frame works in Qualitative Research. Thousand Oaks, CA: Sage.
Boss, P. G., Doherty, W. J., LaRossa, R, Schumm, W. R. and
Steinmetz, S. K., (Eds.) (1993). Sourcebook of Family Theories
and Methods: A Contextual Approach.
Johnson R. B, & Christensen L.B (2007). Educational Research:
Quantitative, Qualitative and Mixed Approaches. NY:
Sage Publications.
Leedy, P. D. & Ormrod, J. E. (2005). Practical Research: Planning
and Design (8th ed,). Pearson Educational International and
Prentice Hall: New Jersey.
McMillan, J.H. & Schumacher S. (2000). Research in Education:
A Conceptual Introduction. London: Longman.
Niss, M. (2006). The Concept and Role of Theory in Mathe matics Education. Paper presented at Norma 05, Trondheim,
2 September 2006.
Tellings, A. (2001). Eclecticism and Integration in Educational
Theories: A Metatheoretical Analysis. Educational Theory,
51, 3, pp. 277-292.
Suppes, F. (1974). The Place of Theory in Educational Research.
Educational Researcher, 3, pp. 3-10.
Oslo, March 2012
Kirsti Klette
Chair of the programme board
7
8
Agnes Tellings, Behavioural Science Institute, Radboud University Nijmegen, The Netherlands
Theories and Research in the Field of Education:
An Indissoluble Union
What is a theory? The word is difficult to define. It is easier to say what a (good) theory does.
It describes reality, explains reality, and, if possible, tells how we can change reality. Thus, it
guides our empirical research and our interventions in practice.
For instance, Piaget’s theory of cognitive development describes four cognitive stages a child goes through, explains
these stages in terms of developing cognitive structures, and
discusses how this development can be helped by the child’s
parents and other educators, that is, by supplying the child
with many different experiences. Familiar as this might seem
to scholars, it is worthwhile to give it some thought once in
a while, for in the daily routines of educational research and
practice one is not always fully aware of the consequences
and implications of this triad between describing, explaining,
and changing reality. An example might illustrate this.
Cambridge Somerville Youth Study was a programme in
which boys from antisocial or otherwise disadvantaged
families were offered different kinds of help, guidance, and
counselling over the period of a few years. Social skills were
taught, healthful and wholesome activities were undertaken,
many features of the riskful environment were changed, and
the boys were given all kinds of medical and social support.
Each boy had one counsellor who coordinated the activities
and functioned as a mentor. The goal was to prevent them
from becoming criminals.
The study was very carefully designed, the counsellors were
trained very well, and next to the group of boys who received
the treatment there was a comparable group of boys from
problematic families who did not receive such treatment.
Thus, it was possible to precisely determine afterwards
whether or not the treatment was effective, and the boys
were followed for many years after in order to also measure
long-term effects.
This intervention was based on former studies which had
shown that boys from at-risk families, living in disorganised
neighbourhoods, had a much greater risk of becoming delinquent than boys who came from good homes (=description of
reality). The theoretical idea was that even boys living in very
bad circumstances could be steered away from delinquent
careers. For this to happen, during their formative years there
should be one adult who could be an example for them and
give the boy emotional support and guidance, and most of
the limiting conditions they lived in should be changed (=explanation of reality and changing it). So, there seemed to be a
sound theoretical idea behind this intervention.
Now, although the boys in the treatment group did show some
positive behavioural changes after the treatment, follow-up
studies showed the boys in the treatment group to have broken
the law significantly more often than the boys in the comparison group who did not receive the same treatment. So the
treatment had reversed effects! An explanation for this phenomenon could only be found in something being wrong with
the theoretical basis of the study. For, as I explained above, the
study was methodologically well executed with well-trained
counsellors. A plausible theoretical explanation indeed has
been given, induced by another finding in the study: that the
boys who had been in summer camp during holidays and had
been treated (this was part of the intervention for most boys)
had broken the law more often, and the more often they had
been in summer camp the higher their crime rate.
According to the construct theory of motivation of McCord
(2000), peers more than parents or other adults shape the behaviour of adolescents – by shaping the motives adolescents
9
have for their behaviour. In these summer camps, children
had a lot of unsupervised time. The boys in the treatment
program probably bragged about their bad behaviour to their
non-criminal peers during summer camps. This gave them a
special status in the peer group, and that motivated them to
stay on the criminal path.
Strangely enough, the detrimental effects of the Cambridge
Somerville youth intervention have not prevented scientists
and policy makers from making the same mistake again and
again. Programmes and empirical studies in which adolescents who are at risk for becoming criminals were obliged to
visit peers who were locked up in prison (so as to deter them
from becoming criminals themselves) have had the same
reverse effects, and policy makers (at least in the Netherlands
where I come from) still advise “education camps” for criminal
youth. These experiences show the need for a good theory. In
this case, what was lacking was a thorough, well-considered
theoretical description and explanation of what influences
and motivates adolescent behaviour.
So, a sound theoretical base for our research and for our interventions is of huge importance. But how do we find, let alone
develop, such a sound theoretical base? There seem to be so
many different theories, standing in different research traditions. For instance, in the field of moral education, character
education and the Kohlbergian approach seem to oppose each
other. To describe the difference somewhat bluntly, according to character education we should teach children moral
behaviour mainly by explicit reward and punishment, whereas
according to Kohlberg’s theory the moral development of the
child progresses more or less automatically via fixed stages,
10
provided that the environment makes sure the child can encounter the right type and diversity of moral experiences. Similarly, in the field of school education the Montessori approach
competes, on the one hand, with more traditional schooling
and, on the other hand, with the so-called “New Learning” in
which the child is entirely free to choose what she wants to
learn and do, provided she does not disturb others.
In order to make a good choice for a theory or an accompanying intervention, one should consider that most theories rest
both on foundations that are very hard to verify empirically
and on empirical research or experience. Which of these two
(the foundation or the experience) comes first, is a chickenegg problem. An example is given below:
“human beings have innate
language structures in their
brain“ (foundation)
theory of Chomsky
“all children learn their
mother tongue in an unbelievably short time without
it explicitly being taught to
them” (experience, empirical research)
«
Theory without practice
is for geniuses, practice
without theory is for
fools and villains, but for
most educators [there]
is the profound, indissoluble union of both
Thus, when choosing a theory to use as a basis for our research
or practice, we must find its foundation(s) at least plausible
and check whether the empirical research it rests on is sound.
The plausibility of not (as yet) verifiable assumptions can only
be safeguarded by good reasoning and probably, in the end, by
intuition. As regards sound empirical research, the ideal design
of an empirical study looks something like this (Veerman &
van Yperen, 2007), for educational intervention studies:
• Clearly described intervention or experiment, sound theory
base, with detailed protocol
• Experimental group gets the intervention or experiment
• Control group gets placebo or alternative intervention or
experiment
• Participants at random assigned to a group
• Participants similar on relevant characteristics
• Pre- and post-test
• Adequate statistical analyses
For pure descriptional or other type of studies this list should
be adjusted, but the general idea is clear. As Veerman & Van
Yperen (2007) explain, in education research it is often not
possible to perform studies that answer the ideal design
because practice is not always manageable, but that is no big
problem as long as one is aware of the lower scientific status
of the results of a less ideal design. It does not at all mean
that the study is valueless. Thus, if we have checked both the
plausibility of the theory’s foundations and the soundness of
the empirical research it rests on, we can make a judgement as
to the general quality of the theory at hand. We then have to
analyse whether the theory is plausible, consistent, coherent,
and well-reasoned.
It is easiest when, in this way, we can deal with the different available theories and end up choosing one particular
theory for our further research and interventions. However,
sometimes it will be better to combine theories. In the field of
education, this is possible because the human beings who are
the centre of education are very complex beings, and theories
that seemingly focus on one and the same aspect of human
behaviour or human functioning, on closer view often turn
out to deal with slightly different aspects that can or should
be combined. Combining or integrating several different theories can roughly be done in four ways (Tellings, 2001):
• reduction
• synthesis
• horizontal addition
• vertical addition
Reduction means that one theory is re-defined in terms of the
other theory or is subsumed under another theory. It is said,
more or less, “If you really look closely to these two theories,
with some corrections and adjustments, they amount to one
and the same core theory.” Or it is concluded “...on closer view,
the one theory is really part of the other theory”.
A well-known example of reduction of terms and, with that,
reduction of theoretical approaches, is the reduction of mind
to matter: the idea that “the mind” is merely a term for a
non-existing something and that all we really have is the cells
in our brain. Although this reduction solves some problems, it
also gives birth to many new ones.
Reduction can also be a first step towards one of the other
three types of integration of theories. For, if we want to
11
«
theory without practice is empty,
practice without theory is blind
integrate different theories or models, we must first be sure
whether, and to what extent, they are really different.
• If they are not really different, we can reduce the one
theory to the other.
• If they are really different, perhaps they can be combined in
one way or the other.
It is as if we have to do some cleaning up first by means of
reduction, before the Real Stuff of combining or integrating
can start.
The second integration type is synthesis. In this type of
integration, the integration of theories leads to entirely new
insights. The theories fertilise each other, new ideas originate
where the two theories or models meet. An example is CAT,
cognitive analytic therapy, designed by Anthony Ryle. This
therapy combines ideas from Freudian analytic therapy with
cognitivism and with ideas from Vygotsky. From Freud it takes
the delving into the psyche of the patient but it leaves aside
the unstructured character of Freudian therapy. From cognitivism it takes the structured, goal-directedness but leaves
aside its superficiality. And from Vygotsky it takes the clear
role of the environment, more particularly of a mentoring
adult – in this case the psychotherapist.
The result of this integration is something new in the sense
that elements are central in the therapy that does not occur
in Freud, or in cognitivism, or in Vygotsky. One such element
is formulation. This takes the form of reformulation of the
problems of the patient early in the therapy process, and of
formulating both unhelpful and helpful procedures for dealing with problems, later on in the therapy process. In these
12
formulations, Freudian insight into the psyche, cognitivist
rational thinking, and the Vygotskyan experience of the mentor, the therapist, come together.
The third integration type is horizontal addition. This type of
integration is useful when different theories or models cover
different domains in education – or when they cover different
aspects of one domain, as for instance empathy development
and cognitive reasoning development, which might be viewed
as two different aspects of moral development. Such different
theories and models, then, are rather simply added together
so that a complete picture originates froms the domain, with
all relevant aspects. The underlying idea of horizontal addition
is that the phenomena described by the different theories or
models occur more or less diachronically, in the same crosssection of time. For the domain of education, with its multidisciplinary character, horizontal addition is often useful.
An example is mathematics. In order to solve fractions, several different skills are required. First, one has to have a concept
of what a fraction is. Mathematical insight, but also language
knowledge, is required for this, and both are acquired by lots
of experience with quantities in the real world. Furthermore,
one must have stored in mind arithmetic facts, such as the
multiplication tables. This can be achieved by rote learning,
which requires a good working short-term memory and longterm memory. These are necessary also for another aspect
of fractions: one must know and understand procedures for
solving fractions. Thus, language, conceptual knowledge,
real-life experience with quantities, short-term and longterm memory, rote learning, and insight play a part in solving
fractions. A good theory and method concerning fractions
must account for all of this and it is likely that a combination
of different theories will work best. Rote learning, which can
be stimulated with rewards, can perhaps best be approached
behaviouristically, whereas insight into what a fraction is can
probably be explained best by cognitivistic theories.
Finally, there is vertical addition. In this form of integration
theories or models are piled on top of each other. The underlying idea of vertical addition is that different theories or models
describe different stages or phases in a development. This form
of integration is especially suitable in the context of education
since children by definition are developing human beings.
Vertical integration has been attempted in moral education
by different scholars, of which Richard Peters was the first.
The idea is that character education and the Kohlbergian approach, seemingly conflicting approaches to moral education,
can be fruitfully combined vertically. Character education says
that morality entails the transfer of moral norms and values
to children with behaviouristic methods, with reward and
punishment. Kohlberg, in contrast, thinks that moral education, or rather moral development, means the development of
choosing one’s own norms and values in a principled and wellconsidered way. Kohlberg’s approach is based on cognitivism.
The vertical integration entails that character education is a
necessary step towards later choosing of one’s own norms and
values. The idea is that one cannot consciously choose certain
norms and values without at least having to some degree internalised these norms and values, by behaviouristic methods.
It is interesting that here not only theories and methods
are integrated, but also foundations. The behaviourism of
character education is viewed as a necessary step towards the
cognitivism of Kohlberg’s moral reasoning.
A final remark with respect to the integration or combining of
theories is that most of what I have said above also holds true
for integrating of combining practical interventions or (school)
methods, which some of the examples I have already given
show. Going back once more to theory as a triad of describing, explaining, and changing of reality, I would like to say
something about both the role of foundations and of empirical
research as pillars of a good theory. Next to foundations that
have as yet not been verified empirically – such as the idea
that language is innate or the idea that children of all ages
are eager to learn by themselves without external force – but
which perhaps once might be verified, there are other views involved in educational theories that can never be verified. They
are the educational norms and values by which we are guided,
more particularly our ideal of the well-educated adult human
being. For instance, Kohlberg’s theory of cognitive reasoning is
based on a very different ideal, that is, of a much more autonomously thinking adult than the character education approach.
When choosing a theory, or when integrating theories, we
must be aware that any choice of theory at the same time
means a choice of one or more educational values.
Usually, researchers do not explicitly choose a theory to base
their research on. They work within the framework of the
theory they happen to be educated in at the university, or the
theory they encountered in the position they happen to apply
for. Nevertheless, of course, in retrospect it is also quite possible to reflect on the theoretical tradition one works in. This
13
should perhaps be done much more, in open discussion with
other traditions.
As regards empirical research, this is the indispensable tool for
developing, verifying and (re)adjusting theories. It can corroborate or falsify theories, or even bring to life new theories. The
relation, however, is mutual. Good theories yield many interesting issues for empirical research. Doing empirical research
without a firm theoretical basis is not only lazy research, it can
have detrimental effects, as we have seen in the case of the
Cambridge Somerville Youth Study and in several successive
interventions and studies with similar approaches. The originators of that first study (almost 70 years ago now!) are not
much to blame. They had a theoretical rationale which was
plausible and which they thought to be sufficient. However,
there should be something like advancing insight, and later
researchers should have learned from the Somerville study.
I would like to end with a quote about the mutual relationship between theory and practice. Many such quotes exist,
most of them having the structure of the one attributed to
the philosopher Immanuel Kant (which, many say, Kant never
uttered): “theory without practice is empty, practice without
theory is blind”. A nice, and for education relevant, one is
the following, which is a quote from the Dutch pedagogue J.
H. Gunning (1859-1951), cited by one of his successors, M.J.
Langeveld (1905-1989; 1979, p. 23): “Theory without practice
is for geniuses, practice without theory is for fools and villains,
but for most educators [there] is the profound, indissoluble
union of both.”
14
References:
Langeveld, M.J. (1979). Beknopte ‘theoretische pedagogiek’.
Groningen: Wolters-Noordhoff.
McCord, J. (2000). A Theory of Motivation and the Life Course.
In Karstedt, S. & Kai-D Bussmann (Eds.), Social Dynamics of
Crime and Control: New Theories for a World in Transition.
Portland, Oregon: Hart Publishing, 229-241.
McCord, J. (2007). Crime and Family: Selected Essays of Joan
McCord. Philadelphia: Temple University Press.
Tellings, A. (2001). Eclecticism and Integration in Educational
Theories: A Meta-theoretical Analysis. Educational Theory,
51 (3), 277-292.
Veerman, J.W. & Van Yperen, T.A. (2007). Degrees of Freedom
and Degrees of Certainty. A Developmental Model for the
Establishment of Evidence-based Youth Care. Evaluation
and Program Planning, 30, 212-221.
15
16
Lars Mjøset, Department of Sociology and Human Geography, University of Oslo
Many Notions of Theory – Too Few
Methodologies to Deal with Them
We can distinguish three ways in which social science methodologists conceive of science.
This distinction is grounded in a fairly detailed sociology of science analysis (Mjøset 2009),
one that we here have to take for granted. The standard view relies on a methods community with the natural sciences (via mathematics and statistics), the social-philosophical view
relies on a symmetric methods community with the humanities (via philosophy, linguistics
and other disciplines based on the interpretation of texts), while the contextualist view relies
only on a methods community internal to social science.
Methodologists define the three methodological views (or
practical philosophies of social science) when they legitimate
their research experiences (based on their skills in one out of
the three clusters of methods) by borrowing concepts and arguments from the relevant areas within professional philosophy of (natural, human and social) science.
is a taxonomy of understandings of science and theory in
present-day social science. The four right-hand columns give
some historical clues. In the following, we survey (however
briefly) both the historical roots and the contemporary notion.
Within each view we find at least two different notions of
theory. In the overview below, we establish a connection between different understandings of theory and the interacting
histories of higher learning/research institutions and the state
in the Western world. Each of the three views connects to a
reference faculty, since faculties unite similar disciplines. The
historical analysis behind these historical clues concerns the
differentiation of the university institution in a setting where
the largest Western states relate both to industrial transformation (impact of basic natural science in the late 19th
century scientific revolution) and to social mobilization. The
detailed processes can be traced in country cases (for Norway,
see Mjøset 2011). In this very short overview we start with the
brief summary contained in Table 1. The first three columns
17
View of science
Reference
faculty
Notion of
theory
Development
phase/university
Historical roots University
institution
Historical roots State/civil
society
Historical examples
Socialphilosophical
Humanities
Reconstructive
Humboldtian
“Bildungs”university
Philosophical (humanities) faculty, dominating the university
Socialization of elites in
pre-democratic states
Encyclopaedic
Hegelian idealist
philosophy
Contextualist
Social science
Critical
(Outside)
Social mobilization of
workers for full political
rights, since the mid-19th
century
Marxism from the
1850s to the 1930s
Standard
Mathematics/Natural
science
Laworiented
Research university
Reference faculty influence on
humanities (and law) faculties:
experiment and statistics
(Second scientific revolution)
All the social sciences,
particularly the most
influential early ones:
psychology, education
and economics
Standard
Mathematics/Natural
science
Idealizing
Research university
Reference faculty influence on
humanities (and law) faculties:
mathematical modelling
(Second scientific revolution)
Economics alone, but
spreading to sociology
and political science
since the 1970s
Contextualist
Social science
Explanationbased
Research university (but partly
outside)
(Social reformist thinking gaining more influence in the 20th
century)
Increasing state activism
since the late 19th century
in social policies, and/or
social reformism in civil
society
Depending on country: anthropology
(Britain), social policy
(Germany).
Socialphilosophical
Humanities
Deconstructive
Mass university
Present day humanities faculty
in the mass university: post
1960s denationalised humanities influence social science
(Socialisation of parts of
the student masses.)
E.g. cultural studies,
media studies.
Table 1. Notions of theory in the social sciences, contemporary taxonomy and historical roots.
18
The social-philosophical reconstructive notion of theory
Historical background: Humboldt turned Kant’s (1798) vision
of a philosophical faculty free to pursue the truth into a postmedieval model of the university. The philosophical faculty
dominated this enlightenment (Bildung) university, trying
all students in preparatory tests on a range of subjects from
classical languages to natural science, synthesised in encyclopaedic systems of Hegelian idealist philosophy. The social task
of the university was to socialise future lawyers, doctors and
priests into national elites. Hardly any new knowledge was
produced. Philosophers lectured on older texts. As research
gradually emerged, it consisted of transcendental reasoning,
philology and case reconstructions, relying on texts (philosophy), archives (history), and the national cultural heritage
(literary texts, cultural artefacts in ethnographic museums).
Except for some field excursions to exotic places, there was
very little observation by participation. The overall notion of
theory was holistic, loosely contextualised and very much preoccupied with the progress through which the human spirit
gets to know itself. This realization, a Hegelian synthesis, was
also the codification of knowledge.
Contemporary notion: A related notion of theory reappears in
some, but not in all the present social sciences. It is distinctly
absent in economics, it is rare also in psychology. We find it
above all in sociology, but it also appears in anthropology,
political science, and education through inspiration from high
sociological theory.
It is a late addition, conquering the curricula only since the
1980s. It was influenced by continental philosophers pursuing “critique of positivism” from the 1960s onwards: most
typically in Habermas’ work (as well as in Giddens, Luhmann,
Alexander and others). There is now a host of secondary introductions to such “social theory” from the sociological classics
and on. General theory amounts to investigating fundamental
preconditions of social science conceived as a set of transcendental concepts (action, interaction, knowledge, and structure). Although they are hardly philosophical idealists, the
social philosophers continue in line with Hegel’s investigation
of mankind’s the present, aiming to specify their fundamental
concepts in grand interpretations of the present state of modernity. In the wholly different context of the mass university
and a democratic political system, their style of theory-making emphasises totality/holism, anti-positivism and the leading role of (continental) philosophy. Such social theorists are
“literary intellectuals”-like analysts who synthesise the spirit
of the age in personal intellectual products. The fundamental
concepts should be general, but there is very little conceptual
convergence between different theorists. It hardly follows a
distinct method other than firstly, interpreting classical texts
of sociology and social philosophy (thus the label “reconstructive”), and secondly, analysing the modern age with reference
to research that has already been carried out. The interpretation of the present may focus on structural features related
to the global or national situation, or individual/existential
personal matters. Such interpretations may be pursued with a
materialistic or an existential orientation.
The contextualist critical notion of theory
Historical background: The historical roots of this notion
are entirely outside of the university institution, as knowledge here is accumulated into theory with reference to
social movements organizing to achieve legitimate social
19
change. Marxism was the first case where a social movement
standpoint claimed distinct merits as social science. In such a
standpoint epistemology, a sociology of knowledge perspective, is linked to ethical considerations. Marx considered the
university economists of mid-19th century Britain as “vulgar
economists” because they produced theories that obscured
the role of the underprivileged workers in economic development. As this was several decades before democratization,
such theories emerged and circulated in activist circles considered by the state to be dangerous enemies. Later, in the more
democratised, but still turbulent, interwar period this impulse
also yielded early reflections on the philosophy of the social
sciences, especially in the philosophical “critical theory” of the
Frankfurt School.
Present notion: Critical theory is an early notion of theory, but
a late one inside of the university. Since the 1970s, Frankfurt
School critical theory influenced the kind of reconstructive social philosophy just surveyed above, becoming integrated into
the university institution. But at the same time, other social
movements gave rise to new, standpoint-based accumulation
of knowledge originating outside of the university. Indications
of critical theory in this sense are theories named after social
movements. The most typical recent case is feminist theory. It
took up the standpoint epistemology tradition pioneered by
Marx. But the context was different (a democratised post-war
society), and so was the focus: the gendered division of labour
not just at work but also at home. Still, the epistemology emphasizing the standpoint of the underprivileged clearly stands
out. Just as Marx saw the whole set of academic economic
theories of his age as “bourgeois”, first generation feminists
saw all academic social science theory in the 1950s and early
20
1960s as “male”. The main argument is that any social science
theory about contemporary developments potentially participate in these social developments, feeding back knowledge to
the actors they are about. This may create or stimulate social
mobilization. Such mobilization, or even just unintended consequences of regular interaction, may change social relations,
make established theories “untrue” and society as a research
topic a moving target (Hacking 1999).
The standard law-oriented notion
Historical background: Around the mid-19th century, separate
faculties of mathematics and natural science broke out of the
philosophical faculties. The two standard notions of theory
emerged inside the university institution, as the new faculties
began to exert influence back onto the humanities faculties
they had broken out of. Through the first half of the 19th
century, natural science introduced systematic observation,
experiment and disciplinary organization into the university.
This became a major source of influence on some fields within
the humanities (and law) faculties, especially the emerging
social sciences. The natural sciences seemed capable of exact
results, rapid progress and practical usefulness. The causal
laws demonstrated by physics experiments (as pioneered by
Galilei) defined a strong ideal. Furthermore, the “assistant”
sciences of mathematics and statistics provided a number
of formalisms, some of which even the social sciences could
try out. The enlightenment university was transformed into
a research university, internationalised through disciplines
(international networks of researchers interacting in research
frontiers), journals and conferences. Entrepreneurs of the new
social sciences offered them as engineering sciences in the
fields of social and economic development. Psychology would
replace Hegelian speculations on the spirit with experimentally based knowledge on perception, memory, reasoning,
etc. Education emerged in close interaction with this kind
of psychology, offering various tests, e.g. of intelligence and
work-related skills. But both the engineering vision and the
strictly law-oriented notion of theory soon created problems.
Social researchers found no laws that compared to the examples they liked to quote from the natural sciences.
tive work is here presented in a terminology originally drawn
from experiments (the effects of independent and intervening variables on dependent variables), but the reasoning is
entirely verbal. Such versions of middle range theories differ
from the contextualist explanation-based theories (discussed
below) only in the philosophy of science reasoning that is borrowed to legitimate the procedures and results.
Present notion: The failure to satisfy such ideals led to notions
such as middle range theories and mechanisms. There were
two varieties of this experience. One was related to statistics.
Within psychology, the technique of randomised controlled
trials spread from pharmaceutical testing of medicines. But
this technique could only demonstrate the effect of a treatment, it gave no clues about the additional causal factors
with which the treatment interacts. This is a problem if the
treatment investigated is not a drug, but for instance a social
policy or therapy intervention. A second experience related to
non-experimental data. Apart from psychology, most social
sciences work with such data, regarding statistical inference as quasi-experiments, reading regression equations as
evidence of imposed experimental controls (the effect of one
factor with all the others kept constant). Econometrics has
developed this approach the furthest. However, promises
that sophisticated formalisms would lead from correlations
to causes failed to deliver (Abbott 2001). These theories
are in any case restricted to the field and the time period
from which the data are collected. Their generalizations are
segmented, and this makes them middle range theories. The
“softest” version of middle range theory appears in parts of
sociology, psychology, education and political science: qualita-
Historical background: As with the law-oriented notion, the
idealizing notion emerged through interaction between the
old and the new faculty. Until recently, this notion was only
relevant in economics. Economics went through two revolutions. First, in the 1870s, mathematical formalisms were
borrowed from the static mechanics (constrained optimization), yielding neoclassical equilibrium theory. Ever since,
economics have been socialised into mathematical problem
solving. Second, in the late interwar period, inferential statistics seemed to offer a way to link the mathematical models
to empirical data. True enough, economics gained status as a
successful social engineering science with its contribution to
the planning of the post-war Western mixed economies. This
was and is a success, but it relied on many empiricist inputs,
and despite it, theory and empirical research did not come
well together. Economic modelling retained the experimental
ideal, since any experiment must provide idealizations in its
efforts to isolate law-like regularities. Although economics
hardly built experiments, the methodologists emphasised
that economic models were thought experiments, and that
their “irrealism of assumptions” (Friedman 1953) was warranted since the models yielded useful predictions.
The standard idealizing notion of theory
21
Present notion: Since the 1970s, sociology and political science
began to borrow the idealised thought experiment (also in its
more sophisticated varieties such as game theory). A fraction
of the economists, however, turned to psychological experimenting (behavioural economics). This implied a connection
to psychology for the first time since economics wedded itself
to mathematical reasoning. This turn has – over some three
decades – introduced many modifications to its ambitious
mathematized equilibrium theory of human interaction in
markets. Among these economists, the idealised notion of
theory is becoming more like yet another collection of middle
range theories. Idealised models never played any prominent
role in psychological theories of action and interaction. Its
dualism is between the law-oriented standard notion and
the contextualist explanation-based notion. Economics, in
contrast, struggles with a dualism between the two standard
notions of theory. Given its methods community with mathematics, the idealised notion tempts scholars in economics
(and some in sociology and political science) to retain the
ideal of knowledge converging into one overarching research
frontier of related, general concepts. The empirical basis is not
primarily the empirical datasets behind law-oriented econometrics findings, but a combination of elementary experiments and a good dose of common sense (what Cartwright
2007: 226, calls “meagre” theory). The main challenge here is
external validity. This has been much discussed also with reference to psychological experiments. But psychology’s laboratory experiments lead on to multiple specific research areas,
while the economists’ models of whole economic systems
are mathematical only. Psychological theories of action have
always been more scattered, being developed in several fields.
22
The contextualist explanation-based notion of theory
Historical background: Critical theory developed as the first
notion that emphasised the participation of knowledge in
social change. This change was first resisted by the state, but
as the labour movement gained strength, a tension developed
between reformist and revolutionary strategies. Projects of
gradual reformist change were pursued by the state (German
social policies were crucial to early German social science)
or by voluntary groups in civil society (the U.S. case). This
interacted with the emergence of experts organised in professions. Their knowledge was predominantly based on experience with intervention in cases. Learning processes are often
based on diffusion of knowledge about exemplary cases. This
style of knowledge accumulation was only partly accepted in
university institutions. However, anthropologists, ethnographers, and sociologists doing fieldwork (e.g. the U.S. Chicago
school of sociology) began to think about such case-based,
bottom up development of theories that retained context
and achieved generalizations that were limited in time and
space. Thus, both critical and explanation-based theories
belong to the contextualist cluster, based on reflection about
how knowledge plays a part in social development.
Present notion: Interest in case studies revived in the 1960s,
even outside anthropology and ethnography. Contributions
such as Glaser & Strauss (1967) suggested a method for case
studies (based on fieldwork particularly) unconnected to any
experimental ideal. Such explanation-based theory is built on
explanations of cases with reference to selected core categories. The contextualist practical philosophy (Mjøset 2009) is
defined as doing what natural science cannot do: enter “into”
the object of study, talking to and interacting with its constit-
uent parts! The two contextualist notions, explanation-based
and critical theory, can be placed as end points on a continuum defined by the nature of participation in the light of
ethical reflection on legitimate strategies of change (reformist
to fundamental). Many combinations in between these two
extremes are possible. Examples are anti-psychiatry, action research, educational experiments. These are not experiments
from which to derive general knowledge. Their purpose is to
achieve change for the better. The researchers are participants. They may have to discuss charges that their intervention was for the worse. Like in critical theory, ethical reflection
is absolutely central here. Knowledge is related to specific
outcomes, as in history, law and therapy. Generalization can
be offered only with context included. If knowledge is diffused, those implementing it must be able to judge their own
context at the receiving end. Reflections on the methodological challenges of qualitative methods such as fieldwork, case
studies, interviewing, participant observation, macro-qualitative comparison, etc. have since the 1960s mainly taken place
in the social sciences. Explanation-based theory may – as in
therapy – be a question of achieving health improvements for
a single client, or a better learning environment for one single
school class. But it may just as well be a question of a larger
unit (state, local government, a private firm, etc.) carrying out
reforms based on knowledge of the state of that unit and its
context. Such case focus is relevant in a wide spectrum of
policy-oriented fields: social work, criminology, schools, care,
etc. It is an important field for employment of social scientists
in the age of interventionist states with large research councils financing problem-oriented research programmes (cf. the
notion of “mode-2” knowledge production; Nowotny, Scott,
Gibbons 2001).
The social philosophical deconstructive notion of theory
Historical background: The youngest notion of theory is another social philosophical notion. The deconstructive notion
resulted from a major post-war transformation of the humanities into post-national humanities. In most countries, the
humanities in the 1950s and 1960s were still part and parcel of
a nation-building project, contributing to the construction of
a national cultural legacy disseminated in schools and media.
But in the late 1960s student revolt, most humanities disciplines came under strong influence from those social science
disciplines that were most influenced by critical theory. Across
the Western world, the humanities disciplines became more
sceptical of their earlier role in sustaining national legacies.
While the reconstructionist social philosopher’s notion reaches
back to pre-research university humanities, poststructuralist
deconstructionism (originating in Paris in the 1960s) attempts
to develop an assessment of the modern world with reference
the most formal and natural science-like part of 20th century
humanities: linguistics. In the history of ideas, art history and
comparative literature, structural linguistics was relied on to
develop theories of discourse and narrative. The French intellectuals were closer to the role of literary intellectuals than in
most other countries. Although often confused with critical
standpoint theorists, poststructuralists actually insisted on
the deconstruction of any standpoint, taking the role of the
ironic artist rather than that of the committed activist. They
implied a highly relativist sociology of knowledge (inspired by
Nietzsche), in which all knowledge is power.
Present notions: – This approach was diffused from Paris to
the Western world via Anglo-American humanities since the
1980s. Consolidated as a reference point especially for the
23
humanities disciplines closest to art, it then trickled down
on the social sciences, bringing with it philosophical borrowings about “construction”. This was an opposite movement
compared to the earlier case of the student revolt, where
social science influenced the humanities. True to its origins in
the humanities, even post-structuralism generates modernity
theories. In addition, however, they produce studies of powerrelations in minute details. Like in critical theory proper, we
are here at the limits of theory. If all accumulated knowledge is power, theory best be avoided. However, in most
cases, scholars pursue poststructuralist theories in academic
projects, retaining the term theory. But since the notion of
deconstruction has intimate links to the arts world, scholars
working with such a notion seem well placed to move into the
cultural-literary public sphere. In that sense, the deconstructive notion of theory fits a society with high living standards
where many young people move to pursue aesthetic pursuit.
Certain disciplines in the mass university provide them with
the enlightenment they need to succeed in such ventures.
But this notion of theory also influenced second and further
generation feminist theorists. Given the victories of the first
generation’s activist critical theories, the second generation could easily move into the university institution. The
poststructuralists’ sociology of knowledge allowed them to
unmask the power of the “malestream” in the social sciences.
Discussing the “constructed” nature of gender, feminist
deconstructionists took the nature/culture dualism quite far.
It is thus not surprising that a response in terms of neo-evolutionary theory, developing in the interface between linguistics
and evolutionary psychology, recently emerged, firmly based
on the standard approach.
How to avoid the methodologists’ escape into philosophy?:
The small survey above adds a university/state-perspective
to earlier efforts at mapping the variety of theory-notions in
the social sciences (Mjøset 2009). The most important finding
is that four notions of theory emerge inside of the university
institution, while two emerge outside of it. The inside notions of theory are connected to tensions emerging from the
internal differentiation of the original philosophical faculty
into natural science and humanities faculties. The two socialphilosophical notions of theory have roots in the oldest and
most recent incarnations of the humanities, while the two
standard notions are return influences by the mathemathical/
natural science faculty on the philosophical faculty. In their
attitude to society, all these four notions relate to the world
from a vantage point within the university institution.
24
In contrast, the two outside – contextualist – notions originally emerged outside of or only on the fringes of the university
institution. They had some influence before, but major influence only after the founding of social science faculties in the
early and middle post-war periods. Inside these faculties since
the 1950s/60s, they represent a legacy from and a link to
knowledge-based reforms and mobilization strategies outside
of the university! Their vantage point is thus not just inside of
the university.
In our accounts of the contemporary versions of the various
notions, we can discern a trend towards stronger emphasis
on middle level notions of theory. While the contextualist
approach only yields theories that are grounded, only valid for
specified contexts, even in the standard and social-philosophical approaches we find a growing interest in such notions. We
shall relate this to another trend: disciplinary commitments
are sustained within the mass university. Our account above
implies that high level notions of theory have a stronger basis
in the university sector than in applied research outside of the
university. The reason is that only the inside views (standard
and social philosophical) include such high level notions. If we
then ask where we find resistance against the trend towards
middle level notions of theory, the answer is plain: primarily at
the universities.
The role of the methodologist – a social scientist with some
specified empirical experience who is not a professional
philosopher – is crucial for this resistance. It may take many
forms. Here we only have space to point to one of its manifestations: the escape into philosophy. University social philosophers are most inclined to escape into continental philosophy,
while standard methodologists tend to escape into analytical
philosophy (or into the connected field of purely mathematical reasoning).
Along these escape routes, methodologists get entangled
with what philosopher Ian Hacking (1999) calls “elevatorwords”: object, truth, reality, objective, construction and the
like. More precisely, they borrow into ontological or epistemological notions – say the distinction between realism and
constructionism, or bold statements on the properties of
various forms of inference – with no reference at all to specific
methods used, specific empirical research results, specific
research problems or research frontiers. Such unspecified borrowing of philosophical terminology is tantamount to “taking
the elevator” to the top floor terrace where the “chill winds of
abstraction” (Leijonhufvud 1975, 334) are blowing. The more
methodologies that are produced in this way, the higher the
risk of stalemate discussions in which scholars “who are not
professional philosophers” fight battles with notions drawn
from high level theories of one or the other type.
Since such battles lead nowhere, it seems important to avoid
them. But discarding philosophy is no solution. That would in
itself be a philosophical position. There may be several productive ways out of this muddle. We shall here focus on one.
Generally, it is relevant for our understanding of theory that we
engage in serious interdisciplinary philosophy, history, and sociology of science. This is what Hacking does in his philosophical
plea to transcend science wars, such as the one that erupted in
1996 with Sokal’s hoax against the poststructuralists. Taken in
isolation, Hacking’s irony towards elevator words might seem
like anti-philosophy. But he is not a philosopher who wants
to tear down all philosophy. He rather wants to deal with the
deepest philosophical challenges by making sense of elevator
words with reference to specific styles of reasoning. True to his
point of departure in analytical philosophy of the natural sciences, Hacking has mainly contributed analysis of the styles implied in natural science (Hacking 2002). But there is an emerging subfield in the borderland between philosophy, history and
sociology of science: “historical epistemology”. One may hope
that research here will provide specifications of styles that are
relevant also to the social sciences and the humanities.
References
Abbott, A. 2001. The Causal Devolution. Pp. 97-127 in Abbott,
Time Matters. Chicago: University of Chicago Press.
Cartwright, N. 2007. Hunting Causes and Using Them.
Cambridge: Cambridge UP.
Friedman, M. 1953. The Methodology of Positive Economics.
Pp. 3-43 in Friedman, Essays in Positive Economics. Chicago:
University of Chicago Press.
Glaser, B. G. & Strauss, A. L. 1967. The Discovery of Grounded
Theory. New York: Aldine de Gruyter.
Hacking, I. 1999. The Social Construction of What? Cambridge,
Mass.: Harvard U.P.
Hacking, I. 2002. Historical Ontology. Cambridge, Mass.:
Harvard U.P.
Leijonhufvud, A. 1975. Life Among the Econ. Western
Economic Journal, 11(3), 327-337.
Mjøset, L. 2009. The Contextualist Approach to Social Science
Methodology. Pp. 39-68 in D. Byrne & C. Ragin, editors,
The Sage Handbook of Case-based Methods, London: Sage.
Mjøset, L. 2011. Samfunnsfagene, universitetet og viten skapen i Norge. Et vitenskapssosiologisk perspektiv på norsk
tenkning og forskning om samfunnet gjennom to hundre
år. Pp. 84-123 in Eivind Balsvik/Susanna Solli, editors,
Introduksjon til samfunnsvitenskapene – en antologi. Vol. 1,
Oslo: Universitetsforlaget.
Nowotny, H., P. Scott & M. Gibbons, 2001. Re-Thinking Science.
Cambridge: Polity.
It should be emphasised that our three practical philosophies
of social science are more aggregated than Hacking’s styles of
reasoning. For instance, the standard approach excels in both
mathematical, statistical and algorithmic styles, while the
contextualist approach excels at least in typological, historicalreconstructive (process tracing), and therapeutic styles. Still, we
shall claim here that a specification of the fundamental problems in the philosophy of the social sciences should as a minimum relate to the three different practical philosophies (view
son science). They have, as shown, all contributed understandings of theory that are currently in use in the social sciences,
being absorbed in different ways by the disciplines. Methodological reflection should thus not relate to social science in an
abstract, aggregated way. Nor should it be related to each of
the single social science disciplines separately. These are harsh
requirements, since they imply that by today, we have very few
acceptable methodologies to go by in the social sciences.
25
26
Siv Haugan, The Research Council of Norway
Multifaceted or one-dimensional?
Education researchers have different ideas about the theoretical challenges of their field.
Here are four researchers’ reflections on the use of theory.
”Research on education builds on many different research
interests. In some areas Norwegian researchers are on the
cutting edge of international research, while in other areas
they lag far behind,” says Professor Peder Haug of Volda University College.
is why the goal of some research today is to say something
about why things are done in one way as opposed to another.
Underlying this is also an element of regulation, where schools
need to document for the governing authorities that they have
done what they are required to do.”
He says that the field of educational theory is highly fragmented and non-homogeneous. ”This can be problematic for
those who are interested in the big picture. For those who
are most concerned with their own topics, it probably doesn’t
matter as much,” says Dr Haug.
Dr Haug emphasises that in studies of relations between
humans, such as in education, it is not possible to predicate
results in the same way as with economic research questions.
Differences perpetuated
The traditional differences in the field of educational theory
still exist:
”On the one hand”, continues Haug, ”we have researchers
who use theory to predicate and regulate, and on the other
hand we have those who use theory mainly to understand
and explain. These ’camps’ have a hard time talking to each
other. Now we’re seeing a growing interest again in predication and regulation.”
Despite this, he believes that the peaceful co-existence of these
contrasting orientations strengthens educational research.
”The result is greater diversity and a better foundation for
knowledge development than if just one of these approaches
was in play.”
”The requirement that schools document their results has
been reinstated after a long absence. The field is much more
concerned with what works than was previously the case. This
Tendency towards one-dimensionality
”A large number of research projects embrace and support a
theoretical perspective based on social constructivism. When
so many researchers employ this perspective, it leads to uniformity. We are seeing a tendency towards one-dimensional
theorising,” Dr Haug believes.
”To shed light on complex issues, such as how learning takes
place in the classroom, a one-dimensional theoretical approach is not sufficient.”
”In order to explain and understand that happens within the
field of education, we probably need to be eclectic. We need
to focus on the many small theories and build them up around
the empirical facts we have. We can’t bring the large theories
down to the classroom.”
Dr Haug emphasises that the point is not which theoretical
foundation gets the ”monopoly”, but that monopolisation is
taking place.
27
There are many thoughts on the role of theory in educational research: (from left) Ivar Frønes, Hans Bonesrønning, Nicoline Frølich and Peder Haug. (Photo: Siv Haugan)
”The result is that many researchers are marching too much
to the beat of the same drum. The critical element in research
can be undermined. Monopolisation also stands in the way of
an eclectic theoretical orientation based on various elements
from a variety of approaches. In other words, a diversity of
theoretical approaches is needed,” Dr Haug emphasises.
Many ways to use theory
Nicoline Frølich, a Research Professor at the Nordic Institute
for Studies in Innovation, Research and Education (NIFU),
believes it is important to ask whether respect for theoretical diversity prevents the accumulation of knowledge across
subject areas and disciplines.
”Education is a complex area of society. There are many different players: teachers, pupils, parents, bureaucrats, politicians.
Many types of processes are involved: teaching, research,
management, administration and cooperation between players. Additionally, there are various types of institutions within
the field: early childhood education and care institutions,
schools, universities, university colleges, vocational schools,
companies, ministries and other public players.”
According to Dr Frølich, this complexity explains why educational research should incorporate insights from many different subject areas and disciplines.
”To exaggerate this, one could say there is no way around
theoretical diversity in educational research,” she says.
”The use of theory is in itself a multifaceted concept, and
there are several different ways theory can be used. It may
28
have to do with looking for patterns, as in natural science.
Theory can also be used the way the classical economists use
it, through deductive approaches of models.”
Who can contribute what?
Hans Bonesrønning, an Associate Professor at the Norwegian
School of Science and Technology (NTNU) and the BI Norwegian Business School, has launched a large-scale project on
administration and management in the education sector.
”We’re studying why the mass educational system functions
so poorly in Norway and what we can do to get more out of
all the years of education. To answer these questions we need
both relevant theory and reliable empirical analyses.”
”Our project is based on common economic decision theory.
But we see that many subject areas can have vital input here,
beyond economics. As economists, we have relatively stylised
models, and we are interested in what psychologists, educators and sociologists can contribute to our work.”
According to Dr Bonesrønning, economists have made some
progress and begun to incorporate elements from other subject areas. They have long used administrative data and data
collected for other purposes in their empirical research.
”If economic research on schools is to advance, a variety
of data will be needed – primarily about teacher and pupil
behaviour, but also about the behaviour of administrators and
behaviour within families. This data capture must be guided
by theory, often by theory from other disciplines.”
«
there is no way around
theoretical diversity in
educational research
”To gain insight into the complexities of the school system,
it is important to have a core model that can be enriched by
perspectives from other subject areas.”
”We must translate theory into reliable empirical research.
Thanks to new data tools, we have accumulated a large
amount of quantitative data, and there is a great need to
analyse this. Knowledge of quantitative methods is needed
to perform these analyses, but case studies should also be
included to flesh things out. But if the ’skeleton’ (quantitative
analysis) is not in place, everything will collapse,” explains Dr
Bonesrønning.
Economic theory can give guidance
Dr Bonesrønning believes that economic theory can be applied in some parts of educational research.
”It’s relatively common for economists to attempt to understand the behaviour of players in the education sector by
applying general economic decision theory, which views the
players as making their choices based on their own preferences and the limitations on resources.”
”One example of this is theories for explaining pupil behaviour.
A common point of departure is to view the pupils’ efforts at
schoolwork as a balance between time for schoolwork and
time for leisure activities. With this as a basis, discussions focus
on, for example, how teachers can influence the pupil’s own
efforts and how the pupil’s efforts are influenced by incentives.
Economic theory can serve as a guide for empirical analyses
that focus on the significance of the students’ own efforts and
various incentives in the school and the labour market.”
”Another example is analyses of the interaction between
players. Some models assume that knowledge production is
dependent on both pupils and teachers make an enormous
effort. Both parties understand that they benefit when the
other party takes the greatest share of the burden. In this case,
a game-like situation arises in which the pupil and teacher
exhibit strategic behaviour in relation to each other. These are
positive theories that can be used to understand how classroom settings actually work,” Dr Bonesrønning emphasises.
”Fashion of the nonsense”
Ivar Frønes, a Professor of sociology at the University of Oslo,
says that theory is crucial for understanding various factors
and for having an overall platform.
”Theory is theory about something. It must be context oriented. When we are going to formulate a theory, we need to
obtain input from other studies of the topic we’re studying.
What previous knowledge do we have about this field?”
”For instance, when we study the problem of marginalisation
in Norway, we must explore the types of contexts and life
courses this involves. We are forced to take a cross-disciplinary
approach, and cross-disciplinarity is necessary for theory
building,” says Dr Frønes.
He notes that researchers should be on guard for ”fashion of
the nonsense”, meaning that researchers become engrossed
with major theoreticians who are in vogue at the time, and try to
make reality fit with their theories.
29
«
We need to focus on the many small
theories and build them up around
the empirical facts we have. We can’t
bring the large theories down to the
classroom level
Dr Frønes says that validity is the critical factor in applied research.
”The desire to conduct applied research that will be used to formulate knowledge-policy naturally puts the focus on validity. The concept of evidence-based measures, and the discussions surrounding
this, illustrates the problem of validity. Measures and policymaking
must be based on reasonably sound knowledge.”
With regard to actions and measures, validity entails more than
showing that the correlation between limited variables is significant. An educational strategy must be grounded in an evaluation
of strategies – in knowledge-based policy, evaluation becomes
all-important.
It is important to evaluate the impacts of educational measures
and strategies. Such evaluations require clear objectives. There are
also limitations on validity since measures have various effects on
different groups and in different contexts. When strategies are to
be implemented, we need to be aware of their potential limitations for certain areas.
30
This publication may be ordered at
www.forskningsradet.no/publikasjoner
The Research Council of Norway
P.O.Box 2700 St. Hanshaugen
NO–0131 OSLO
Telephone +47 22 03 70 00
Telefax +47 22 03 70 01
[email protected]
www.rcn.no/english
Editor:
© The Research Council of Norway,
UTDANNING2020
www.rcn.no/utdanning
March 2012
ISBN 9978-82-12-03050-3 (print)
ISBN 978-82-12-03051-0 (pdf)
Number of copies: 500
Printing: 07 Gruppen as
Design: Jentestreker as
Photo/illustration: Shutterstock